So the only thing I have to add is this: there are not many parts of the circuit that can cause the symptoms you have.
The CPU reads the fwd/reflected power from those two points in the schematic below T507.? Something is wrong with the signal getting there.
That signal must go through T507 and the bridge circuit below it (which is unlikely to have a failure, unless one of the components was damaged during assembly.
To get to T507, the signal must go via C526, a simple .01 coupling capacitor (again, unlikely to fail unless disturbed by soldering during assembly.
And to get there, the signal must come from the selected one of the LPFs via its two diodes.
Since you believe there is no problem with T507, and since your 20M T results are a bit different than the other bands, perhaps we could deduce that the 20M filter has a fault, perhaps causing it to be always partially enabled, causing issues when any of the other filters is also enabled.? This would happen if one of its diodes were shorted (D511, D508).? These can fail.? Or perhaps the switch that turns on the LPF is failed 'on' - this would cause anomalous behavior on the other bands (Q509).
The way I read things, there's not much else that could cause what you are seeing, unless you have several simultaneous failures, which is unlikely.
So that's all I have.
Except one final note: in a previous thread last month, someone was sure his transformer was wound correctly and properly installed.? When all else failed to fix his problem, he finally removed the transformer that he was sure had no problem, rewound it and re-installed it, and everything started working.? So sometimes the error is where we are sure it is not....
Good luck
Stan